UNIVERSITÀ DI PISA 1 34 3 IN S U PR EMÆ DIGN IT A T IS 1
GRADUATE COURSE IN PHYSICS
2UNIVERSITY OF PISA
3
Graduate School of Basic Sciences
4 “GALILEO GALILEI” 5 XXIII ciclo 2007-2011 6
PhD
THESIS
7Measurement of the multi-jet
8cross-sections with the ATLAS
9detector at the LHC
10Candidate
Advisor
11
Zinonas Zinonos
Prof. Vincenzo Cavasinni
PhD Advisor :
Prof. Vincenzo Cavasinni
- University of Pisa
- & INFN Pisa
PhD Committee
President :
Prof. Kenichi Konishi
University of Pisa
Examinators :
Prof. Peter Loch
- University of Arizona
Prof. Elisabetta Barberio
- University of Melbourne
Prof. Franco Bedeschi
- INFN, Sezione di Pisa
Prof. Mauro Dell’Orso
- University of Pisa
Prof. Gavin Salam
- CERN
Dedicated to my wonderful wife
15
Louiza
Contents
Acknowledgements v 18 Preface vii 19 1 Introduction 1 202 The Large Hadron Collider 7
21
2.1 A Short Introduction to Lhc . . . 7
22
2.2 Impediments to High Luminosity . . . 10
23
3 Overview of the ATLAS Detector 13
24
3.1 Kinematic Definitions . . . 14
25
3.2 ATLAS Magnets and Magnetic Field . . . 16
26 3.2.1 Central Solenoid . . . 17 27 3.2.2 Toroid . . . 17 28 3.3 Tracking System . . . 18 29
3.3.1 Overview of the ATLAS Inner Detector . . . 18
30
3.3.2 Pixel Detector . . . 19
31
3.3.3 Semi-Conductor Tracker . . . 20
32
3.3.4 Transition Radiation Tracker . . . 20
33
3.3.5 Readout systems . . . 21
34
3.3.6 Cooling . . . 22
35
3.4 ATLAS Calorimetry Overview . . . 22
36
3.4.1 Overview of the LAr Calorimeter . . . 25
37
3.4.2 Overview of the Tile Calorimeter . . . 27
38
3.4.3 The Forward Calorimeter . . . 29
39
3.4.4 The Zero Degree Calorimeters . . . 29
40
3.4.5 Tile Readout, Reconstruction, Calibration & Performance 30
41 3.5 Muon Spectrometer . . . 34 42 3.5.1 Overview . . . 34 43 3.5.2 Precision-tracking chambers . . . 35 44
3.5.3 Monitored Drift Tube Chambers . . . 37
45
3.5.4 Cathode-Strip Chambers . . . 37
46
3.5.5 Alignment system of the precision chambers . . . 37
47
3.5.6 Trigger Chambers . . . 38
48
3.6 Trigger and Data Acquisition . . . 40
49 3.6.1 L1 Trigger . . . 40 50 3.6.2 L2 Trigger . . . 41 51 i
3.6.3 Event Filter Trigger . . . 42
52
3.6.4 Jet Trigger Algorithm . . . 42
53
3.6.5 Data Acquisition . . . 44
54
3.7 Luminosity Determination . . . 44
55
3.7.1 Basics of Luminosity Measurement . . . 45
56
3.7.2 Luminosity Detectors . . . 47
57
4 Standard Electroweak Theory and QCD 51
58 4.1 Introduction . . . 51 59 4.2 Gauge Invariance . . . 52 60 4.2.1 Quantum Electrodynamics . . . 52 61 4.2.2 Quantum Chromodynamics . . . 54 62 4.3 Electroweak Unification . . . 60 63 4.3.1 Experimental Observations . . . 60 64
4.3.2 The Unbroken SU (2)L⊗ U(2)Y Theory . . . 62
65 4.3.3 Charged-current Interactions . . . 64 66 4.3.4 Neutral-current Interactions . . . 64 67 4.3.5 Gauge self-interactions . . . 66 68
4.4 Spontaneous Symmetry Breaking and Higgs Sector . . . 67
69
4.4.1 Goldstone-Nambu Theorem . . . 68
70
4.4.2 The Higgs Mechanism . . . 69
71
4.4.3 Parameters and Predictions of the Gauge Sector . . . 72
72
4.4.4 The Higgs Boson . . . 73
73
4.4.5 Fermion Masses . . . 73
74
4.4.6 Higgs Production and Decays . . . 74
75
4.4.7 Higgs Mass Bounds . . . 76
76 4.5 Aspects of QCD . . . 77 77 4.5.1 QCD Langrangian . . . 77 78 4.5.2 Why SU (3)C? . . . 78 79
4.5.3 The Renormalization Group Equations . . . 79
80
4.5.4 Running Coupling and Renormalization Scale . . . 79
81 4.6 Structure of QCD Predictions . . . 81 82 4.6.1 Inclusive Cross-Sections . . . 81 83 4.6.2 Scale Dependence . . . 82 84
4.6.3 DIS, PDFs and Factorization Scale . . . 82
85 4.6.4 Hadron Collisions . . . 84 86 4.6.5 Accuracy of Predictions . . . 84 87 4.7 Experimental Observations . . . 85 88 4.7.1 Jets . . . 85 89
4.7.2 Two and Three-Jet Productions . . . 86
90
4.7.3 Strong Coupling Measurements . . . 88
91
5 QCD Multi-jet Measurements 91
92
5.1 Introduction . . . 91
93
5.2 Cross-Section Definitions and Kinematics . . . 92
94
5.3 Theoretical Predictions . . . 95
95
5.3.1 Parton Shower Generators . . . 96
96
5.3.2 Multi-parton Generators . . . 97
97
5.3.3 NLO ME Generators . . . 98
98
5.3.4 MC Generators and Tunes . . . 99
99
5.4.1 Data Statistics and Integrated Luminosity . . . 100 5.4.2 Trigger Performance . . . 101 102 5.4.3 Trigger Selection . . . 105 103 5.4.4 Vertex Reconstruction . . . 107 104
5.4.5 Pile-Up and Jet Vertex Fraction Discriminant . . . 108
105
5.4.6 Jet Reconstruction and Calibration . . . 112
106
5.4.7 Event and Jet Selection Criteria . . . 115
107
5.5 Data Correction . . . 117
108
5.6 Systematic Uncertainties . . . 120
109
5.6.1 Jet Energy Resolution . . . 120
110
5.6.2 Jet Energy Scale Uncertainty . . . 122
111
5.6.3 Systematic Uncertainties in the Luminosity . . . 135
112
5.6.4 Systematic Uncertainties in the Trigger Efficiency . . . . 135
113
5.6.5 Combination of All Systematic Uncertainties on the JES . 135
114
5.7 Systematics on NLO Predictions . . . 138
115
5.7.1 Next-to-leading-order Scale Choice . . . 139
116
5.7.2 Combined PDF, Scale and αsUncertainties . . . 142
117
5.7.3 Non-perturbative QCD Corrections . . . 142
118
5.8 Results . . . 150
119
5.8.1 Jet Inclusive Multiplicity . . . 152
120
5.8.2 Jet Transverse Momenta . . . 152
121
5.8.3 Event HT . . . 157
122
5.8.4 R32 Measurements with R = 0.4 Jets . . . 157
123
5.8.5 R32 Measurements with R = 0.6 Jets . . . 161
124
5.8.6 Angular Distributions . . . 162
125
5.8.7 Probing Parton-Shower Effects . . . 164
126
5.8.8 Multi-jet Cross-sections in Other Physics Channels . . . . 166
127
5.9 Summary and Conclusions . . . 169
128
6 Synopsis 171
129
Appendices 173
130
Appendix A The ATLAS Detector 175
131
Appendix B Multi-jet Analysis 193
132
B.1 Pythia Tunes, PDFs and Normalization . . . 193
133
B.2 ALPGEN and MLM . . . 193
134
B.3 Simulated data samples . . . 196
135
B.4 Trigger . . . 211
136
B.5 Jet Energy Scale and Uncertainty . . . 211
137
B.5.1 Overall multi-JES uncertainty . . . 211
138
B.5.2 Overall JES uncertainty . . . 213
139
B.6 Jet Quality and Jet Cleaning Cuts . . . 213
140
B.7 Debug Stream Events . . . 218
141
Bibliography 243
142
Acknowledgements
I am most indebted to my advisor Vincenzo Cavasinni for his endless help,
144
precious guidance and continuous support. I am particularly grateful for his
145
extraordinary patience and kindness.
146
I thank the rest of the ATLAS Pisa group members, Nino Del Prete,
147
Chiara Roda, Calderini Giovanni, Paola Giannetti, Alberto Annovi, and my
148
colleagues Paolo Francavilla, Michelle Cascella, Federico Bertolucci, Francesco
149
Nuti, Francesco Crescioli and Daniele Puddu. I also owe thanks to the
ex-150
members of the Pisa team Iacopo Vivarelli, Andrea Dotti, Francesca Sarri,
Gi-151
angiobbe Vincent, Roberto Agostino Vitillo, Debenedetti Chiara and Claudia
152
Bertella.
153
I owe thanks to my friends outside of the Pisa team, Elisabetta Barberio,
Pe-154
ter Loch, Vassili Kazanine, Nikiforos Nikiforou, Natalie Heracleous,
Constanti-155
nos Melachrinos, Andreas Petrides, Monika Herodotou, Attikis Alexandros,
Je-156
had Mousa, Christos Hadjivasileiou, Maria Kameri, Maikantis Georgios, Giulio
157
Usai and Claudio Santoni. Their advice, encouragement, ideas and discussions
158
have been invaluable.
159
Special thanks goes to Nektarios Benekos for his exceptional help, inspiration
160
and fruitful discussions.
161
I thank my former advisor Prof. Ptochos Fotios, who first sparked my
inter-162
est in experimental high energy physics and planted my feet firmly on the path
163
to a career in physics.
164
Finally, I cannot forget my wonderful wife Louiza Manoli, whom I thank for
165
supporting me in pursuing a goal that I think cannot be reached without an
166
endless supply of love, understanding and encouragement.
167
Preface
The Large Hadron Collider (Lhc) is the world’s largest and highest-energy
par-169
ticle accelerator designed to accelerate and collide proton beams to the highest
170
energies and luminosities ever achieved in the history of experimental High
En-171
ergy Physics. The Lhc was built to help scientists to answer key unresolved
172
questions in particle physics and to test the known areas of physics at even
173
higher energy regimes.
174
For the past few decades, physicists have been able to describe with
increas-175
ing detail the fundamental particles that make up the nature at microscopic
176
scales and the fundamental interactions between them. This understanding is
177
encapsulated in the Standard Model of particle physics. Developed throughout
178
the mid to late 20th century, the current formulation was finalized in the mid
179
1970s upon experimental confirmation of the existence of quarks. Since then,
180
discoveries of the bottom quark (1977), the Z and W gauge bosons (1983), the
181
top quark (1995), the tau neutrino (2000) and the great success in explaining a
182
wide variety of experimental results have given credence to the Standard Model.
183
Quantum chromodynamics (QCD), an important sector of the Standard
184
Model, is the theoretical framework built to formulate the strong interaction, a
185
fundamental force describing the interactions of the quarks and gluons,
collec-186
tively called partons, making up hadrons. A huge body of experimental evidence
187
for QCD has been gathered over the years in collider physics. Quarks and
glu-188
ons, being produced in large abundance at hadron colliders, like the Lhc, evolve
189
to form experimental signatures in detectors, the jets of the so-called hadronic
190
final state.
191
Jet production and properties are key observables in high-energy particle
192
physics. They have been measured so far in many beam colliders, such as
proton-193
proton, proton-antiproton, electron-positron and electron-proton both at GeV
194
(LEP, HERA) and TeV-energy scales (Tevatron) as well. They have provided
195
precise measurements of the strong coupling constant, have been used to obtain
196
information about the structure of the proton, and have become important tools
197
for understanding the strong interaction and searching for physics within and
198
beyond the Standard Model.
199
Measurements of the jet cross-section and characteristics in proton-proton
200
collisions at an unprecedented center-of-mass energy of 7 TeV at Lhc, were
201
recently performed by the Atlas experiment. The work constituted by this
202
thesis represents a part of these measurements; an endeavor to understand some
203
of the most important experimental aspects of the QCD theory, and in particular
204
the production of multi-jet events in proton-proton collisions.
205
This thesis is devoted to describing the analysis set up to measure the
multi-206
jet production rates and their kinematic features in early data provided by Lhc.
207
A comparison is made between the theoretical predictions and the experimental
208
results obtained by the dedicated analyses performed on the data collected by
209
the Atlas experiment. Special emphasis is laid on understanding the various
210
sources of systematic uncertainties on the experimental results, together with
211
the uncertainties of the theoretical calculations.
212
The author was actively involved in many efforts during the first acquisition
213
of √s = 900 GeV and then √s = 7 TeV proton-proton collision data, and in
214
several jet measurement analyses using the full 2010 dataset.
215
Starting in 2009, he studied the kinematic properties of jets produced in
216
pp collisions at √s = 900 GeV (presented at IFAE2010 and published by SIF
217
in Nuovo Cimento C [1]) and later he participated in the first observation of
218
energetic jets in√s = 7 TeV data collected in 2010 (presented at PLHC2010 [2]).
219
He also made a significant contribution to the first measurement of inclusive jet
220
production cross-section (presented at 35th ICHEP2010 [3], at HCP2010 [4]
221
and published in EPJC [5]) and had a major role in the first measurement
222
of multi-jet production cross-sections (presented at HCP2010 [6]). Finally, in
223
2011, particular emphasis was placed on the measurement of the multi-jet
cross-224
sections with a larger data sample (presented at 46th Rencontres de Moriond
225
on QCD and High Energy Interactions [7], at PisaJet2011 [8] and published in
226
EPJC [9]).
227
The author has participated in all phases of the multi-jet cross-section
anal-228
ysis, from the first measurement to the recent publications, and had many
re-229
sponsibilities during these efforts. Of course, this achievement was only possible
230
thanks to the outstanding competence, dedication and efforts of many people
231
working together in the multi-jet team.
232
Most of the produced results is the outcome of a collaboration, in which
233
the author had a leading role. In particular, he has been direct responsible for
234
the multi-jet trigger performance studies and the trigger design for the event
235
selection. He was responsible for calculating the total integrated luminosity
cor-236
responding to the data periods used in the analysis. He developed the Monte
237
Carlo machinery for generating all multi-jet observables and was responsible for
238
the massive production of simulated events with different leading-order Monte
239
Carlo programs attached to numerous tuning parameters. During the Monte
240
Carlo production, several studies were focused particularly on understanding the
241
differences observed between the simulated samples with different Monte Carlo
242
tunes. Emphasis was put on studying the effects resulting from the choice of the
243
parton distribution functions implemented in the simulation. Among his
respon-244
sibilities, was also the estimation of non-perturbative QCD correction factors
245
needed for the next-to-leading-order theoretical calculation using an enormous
246
amount of simulated events. A lot of contribution was given in understanding
247
the impact of implementing the non-perturbative corrections in the theoretical
248
predictions at next-to-leading-order and their effect on the overall theoretical
249
systematic errors. He participated in the special validation tasks undertaken to
250
control the theoretical calculations at leading-order and next-to-leading-order,
251
to test data unfolding factors, to extract the jet energy scale uncertainty in the
252
data, to understand the impact of pile-up on the measurements and to make
253
estimations of the overall systematic uncertainty bands.
254
He had also a major role in designing the cut-flow of the analysis, in
con-255
structing comparison tables and performing final cross-checks of all final results.
256
This thesis is organized as follows. Chapter 1 provides an introduction and
257
infrastructure at Cern, used to produce, collect, reconstruct and analyze the experimental data. Chapter 2 contains basic information about the Lhc, the
260
machine which accelerates and collides proton beams. Chapter 3 describes the
261
apparatus used to record and reconstruct the proton collision events: the Atlas
262
detector. A basic description is given on the detector’s functionality, both at
263
hardware and software level. Chapter 4 gives a general theoretical introduction
264
in the Standard Model and a brief review on the Higgs production mechanisms
265
in hadron colliders and decays. Also, this chapter provides an phenomenological
266
overview of Quantum Chromodynamics. The description of the main analysis
267
is done in Chapter 5. All small pieces of the analysis are coherently matched
to-268
gether in a big chapter, including relevant theoretical aspects and experimental
269
techniques, leading to the final results. An introduction in the multi-jet analysis
270
is given first in Section 5.1, followed by the definition of the cross-sections to be
271
measured (Section 5.2), a discussion of the simulations used in the measurement
272
and the theoretical predictions to which the data are compared (Section 5.3).
273
Event selection and reconstruction are described in Section 5.4, including
in-274
formation on the trigger performance, the jet reconstruction and pile-up. Data
275
correction is then described in Section 5.5. The evaluation of the main
uncer-276
tainties in the measurement, mainly coming from the jet energy scale, is given
277
is Section 5.6. The systematics in the theoretical predictions are described in
278
Section 5.7, followed by the results and conclusions in Sections 5.8 and 5.9.
279
In Section 5.8.8, the calculated multi-jet cross-sections are compared to other
280
Standard Model physics channels already measured or expected to be measured
281
by Atlas. The thesis is concluded in Chapter 6 by summarizing all findings
282
and making suggestions for future analyses.
283
Chapter 1
284
Introduction
285At the Large Hadron Collider (Lhc), jet production is the dominant high
286
transverse-momentum process providing a direct test of Quantum
Chromody-287
namics (QCD) physics at the TeV scale. In fact, the Lhc physics program with
288
proton-proton (pp) collisions at 7 TeV center-of-mass energy, for the time
pe-289
riod 2010-2012, allows QCD physics to be tested in an entirely new high energy
290
regime.
291
One of the most striking features of Lhc final states is the large number
292
of events with several hard jets. QCD multi-jet events are being produced in
293
large abundance and multi-jet cross-sections are among the first measurements
294
possible to perform with real data collected by the Atlas detector. Lhc is
295
offering thus for the first time the possibility to explore perturbative QCD at
296
the largest center-of-mass energies and probe fundamental interactions at the
297
smallest distances ever achieved in the history of collider physics.
298
A detailed understanding of QCD jet production is very important for the
299
QCD gauge non-abelian theory itself and for all almost the physics processes
300
to be studied at the Lhc. The measurement of jet production cross-sections
301
at Lhc provides a stringent test of perturbative QCD in an unexplored energy
302
regime never probed so far. Extending the kinematic limit of the partonic hard
303
scale, Q2, at the TeV2 order, leading and next-to-leading-order perturbative
304
QCD predictions can be tested with experimental data. The jet cross-section
305
measurements can be thereby used as observables for the determination of the
306
running strong coupling constant up to TeV-scales with high accuracy, as well
307
as show sensitivity to the proton’s parton densities over a wide range of scale
308
and momentum fraction.
309
QCD studies are also relevant for the searches of physics of and beyond
Stan-310
dard Model; new physics. In fact, the multi-jet QCD events are a significant
311
background to many physics channels and being able to measure and describe
312
their features at this high energy regime is essential. The immense amount
313
of available phase-space and the large acceptance of the modern detectors like
314
Atlas and Cms, with calorimeters covering a region of almost 10 units of
pseu-315
dorapidity, can lead to the production and identification of final states with 10
316
or more jets. These kind of events would hide or strongly modify all possible
317
physics signals which involve multi-jet production in the final state, such as
top-318
antitop quark hadronic decays, di-tau lepton hadronic decays and vector boson
319
production in association with jets. Moreover, searches for new physics like
Su-320
persymmetry and new phenomena, typically require very large jet multiplicities
321
and hence they are fully exposed to the high production rate of QCD multi-jet
322
prompt processes. Consequently, a precise knowledge of final states which
in-323
volve significant hadronic activity is required. Precision measurements with pp
324
collision data and accurate predictions for QCD jet event rates (cross-sections
325
and distribution shapes) must be thus performed.
326
Jet cross-sections and properties are key observables in high-energy particle
327
physics. They have been measured at e+e−, ep, p¯p, and pp colliders, as well as
328
in γp and γγ collisions. They have provided precise measurements of the strong
329
coupling constant, have been used to obtain information about the structure
330
of the proton and photon, and have helped to understand better the strong
331
interaction and search for physics beyond the Standard Model. Multi-jet
cross-332
sections measurements have been performed at the Tevatron in p¯p collisions at
333
1.96 TeV center-of-mass energy. Both experiments, Cdf [10, 11] and D0 [12, 13]
334
have measured the multi-jet cross-section, event shapes and invariant mass of
335
systems with up to 4 jets in the final state. Also, the Cms collaboration has
336
recently released measurements of the three-jet to two-jet cross-section ratios at
337
a 7 TeV center-of-mass energy [14].
338
In this thesis, the first measurements of inclusive multi-jet cross-sections
339
using the Atlas detector are presented. These measurements are performed
340
using a data set of pp collisions at 7 TeV center-of-mass energy, taken early in
341
Lhc running in 2010, corresponding to an integrated luminosity of ∼ 2.4 pb−1.
342
The measurement involves a precise determination of the jet trigger and
recon-343
struction efficiencies of Atlas for jets, as well as a first determination of the
344
calorimeter response to jet energy, jet flavor and closest distance to other jets
345
in dense hadronic environments.
346
The Atlas detector is instrumented over almost the entire solid angle around
347
the pp collision point with layers of tracking detectors, electromagnetic and
348
hadronic calorimeters. The calorimeters are surrounded by the muon
spectrom-349
eter which consists of three large superconducting toroids, a system of precision
350
tracking chambers, and detectors for triggering. The inner detector, consisting
351
of silicon pixel and micro-strip detectors as well as a transition radiation tracker,
352
is immersed in a 2 T solenoidal magnetic field. Atlas has a three-level trigger
353
system, with the first level trigger being based on custom-built hardware and
354
the two higher level triggers being realized in software. Jet measurements are
355
made using a finely segmented hermetic calorimetric system, designed to provide
356
three-dimensional reconstruction of particle showers and detect high energy jets
357
with high efficiency and excellent energy resolution up to |η| . 4.9.
358
Individual jets are identified and built using the anti-ktjet algorithm with
359
two jet resolution parameters, R = 0.4 and 0.6. This algorithm is well-motivated
360
since it can be implemented in next-to-leading-order perturbative QCD
calcu-361
lations, is collinear and infrared-safe to all orders and produces geometrically
362
well-defined “cone-like” jets. The inputs to this algorithm are three-dimensional
363
clusters of calorimeter cells with energy depositions significantly above the
mea-364
sured noise. Jet four-momenta are constructed as the vectorial sum of clusters
365
of cells, treating each cluster as a four-vector with zero mass, assuming that the
366
corresponding particle stems from the primary vertex. The jet four-momenta
367
are then corrected as a function of pseudorapidity and transverse energy for
368
various effects, the largest of which are the hadronic shower response and the
369
detector material distribution. This is done using an energy calibration scheme
3 based on Monte Carlo studies including full detector simulation and validated
371
with extensive test-beam and collision data studies.
372
The event selection in the analysis starts with the first-level two and
three-373
jet triggers which collect events that have at least two or three large
trans-374
verse energy depositions in the calorimeters. In the dataset used, the two-jet
375
triggers were prescaled. So, in order to achieve the highest possible effective
376
integrated luminosity, the inclusive two-jet events have been measured by
us-377
ing an exclusive combination of all di-jet triggers available in that data period.
378
The candidate multi-jet events are then required to satisfy certain data quality
379
criteria, to eliminate various detector effects, to suppress beam and other
non-380
collision backgrounds, and have a primary collision vertex defined by multiple
381
charged-particle tracks.
382
Cross-sections are calculated in bins of inclusive jet multiplicity, meaning
383
that an event is counted in a jet multiplicity bin if it contains a number of jets
384
that is equal to or greater than that multiplicity. Inclusive multiplicity bins
385
are basically used because they are stable in the perturbative QCD fixed-order
386
calculation, unlike exclusive bins. Only jets in the kinematic region pT > 60 GeV
387
and |y| < 2.8 are counted in the measurement. These cuts are chosen to ensure
388
that the jets are reconstructed with high efficiency and they are in a kinematic
389
region where the jet energy scale is well understood. The leading jet is further
390
required to pass a higher transverse momentum cut, at 80 GeV, so as to stabilize
391
the next-to-leading-order perturbative QCD calculations in the dijet case.
392
All measured multi-jet cross-sections are corrected for all experimental effects
393
using an unfolding technique with simulated events, to allow comparisons with
394
particle-level theoretical predictions.
395
The theoretical motivation for measuring multi-jet final states is twofold; to
396
evaluate the robustness of the leading-order perturbative QCD calculations in
397
representing the high jet multiplicity events and to test fixed-order perturbative
398
QCD calculations at next-to-leading-order. For the leading-order comparisons,
399
events with up to six jets in the final state are studied. For the
next-to-leading-400
order perturbative QCD study, the focus is on three-jet events and their
com-401
parison to two-jet events.
402
Many different effects are included in leading-order Monte Carlo simulations
403
of jet production at the Lhc. These include the modeling of the underlying
404
event and hadronization which can affect the cross-section calculation through
405
their impact on the jet kinematics. Effects arising from differences between
406
the matrix-element (with up to 2 → 6 matrix-element scattering diagrams)
407
plus parton-shower calculation and the parton-shower calculation alone (with
408
only 2 → 2 matrix-element scattering diagrams) also need to be understood.
409
These topics are not easily separable, since tuning of some of the effects, such
410
as the underlying event, to data is needed. The tuning process, however,
auto-411
matically fixes other inputs in the Monte Carlo simulation, such as the parton
412
distribution functions, the parton-shower model and the hadronization model.
413
The non-triviality to perfectly separate out some effects introduces a difficulty
414
in obtaining a full estimate of the theoretical uncertainty associated with the
415
leading-order Monte Carlo predictions. Therefore, one of the aims of this
anal-416
ysis is to test the performance of the different leading-order Monte Carlo
sim-417
ulations, so that they can be used to estimate multi-jet backgrounds for new
418
particle searches, not to discern whether deviations with respect to QCD are
419
present in the data. The latter goal is best achieved by comparing with
to-leading-order perturbative QCD calculations.
421
The next-to-leading perturbative QCD calculation used in this analysis, is
422
not interfaced to a Monte Carlo simulation with hadronization and other
non-423
perturbative effects and hence making practically the predicted partonic
cross-424
sections unmeasurable. For comparison with data at the particle-level
simula-425
tion, soft (non-perturbative) corrections must be applied to the next-to-leading
426
perturbative QCD calculation. This is done using leading-logarithmic parton
427
shower Monte Carlo programs in which the soft QCD effects are enabled and
428
disabled, and then evaluating the relative ratio.
429
For high multiplicity studies, which include events with up to six jets, the
430
resolution parameter in the jet reconstruction is fixed to R = 0.4 to contend with
431
the limited phase-space and to reduce the impact of the underlying event in the
432
jet energy determination. For testing the next-to-leading-order perturbative
433
QCD calculations, where the study focuses on three-jet events, a resolution
434
parameter of R = 0.6 is preferred, since a larger value of R is found to be less
435
sensitive to theoretical scale uncertainties.
436
The cross-section as a function of the inclusive jet multiplicity of up to six
437
jets is measured. A study that substantially reduces the impact of systematic
438
uncertainties is the ratio of the N -jet to (N − 1)-jet cross-section as a function
439
of multiplicity. In this ratio, the impact of the jet energy scale uncertainty is
440
significantly reduced.
441
Multi-jet kinematic features are also explored. Differential cross-sections for
442
multi-jet production are measured as a function of transverse momentum of the
443
leading, second leading, third leading and fourth leading jet, and their scalar
444
sum as well.
445
A measurement with particular sensitivity to limitations in the leading-order
446
Monte Carlo simulations and next-to-leading-order perturbative QCD
calcula-447
tions is the ratio of the inclusive three-to-two-jet differential cross-section as a
448
function of different kinematic variables. In this measurement, uncertainties in
449
the jet energy scale are significantly suppressed. The three-to-two-jet ratio as a
450
function of the leading jet transverse momentum can be used to tune the Monte
451
Carlo simulations for effects due to final state radiation. The three-to-two-jet
452
ratio as a function of the transverse momentum scalar sum of two leading jets,
453
is found to give the smallest theoretical scale uncertainty and is, therefore, most
454
sensitive to input parameters such as the strong coupling constant αS.
455
A dedicated study of angular distributions in inclusive three-jet systems is
456
also performed. Polar and azimuthal differences between the first three leading
457
jets are measured in data and compared to simulation using different leading
458
and next-to-leading-order Monte Carlo programs.
459
Special emphasis is given on the evaluation of the jet energy scale and its
460
uncertainty, which is the dominant uncertainty source for most results presented
461
in this analysis. The fact that cross-sections are steeply falling as a function
462
of jet transverse-momentum implies that, even a relatively small uncertainty in
463
the determination of the jet transverse-momentum translates into a substantial
464
change in the cross-sections as events may migrate along the steeply falling
465
curve.
466
The jet energy scale and its uncertainty have been determined for jets from a
467
dijet sample without nearby activity in the calorimeter. For a multi-jet analysis,
468
additional systematic uncertainties need to be considered. These uncertainties
469
arise from the difference in the calorimeter response to jets of different flavors
5 as well as the impact of the presence of nearby activity in the calorimeter on
471
the jet energy measurement.
472
For events containing two or more jets, a reasonable agreement is found
473
between data and leading-order Monte Carlo simulations with parton-shower
474
tunes that describe adequately the Atlas√s = 7 TeV underlying event data.
475
The agreement is found after the predictions of the Monte Carlo simulations are
476
normalized to the measured inclusive two-jet cross-section.
477
All models reproduce the main features of the multi-jet data. The 2 → 2
478
QCD calculations show some departure from the data for the three-to-two jet
479
cross-section ratios, predicting a higher ratio than observed. The 2 → n ≤ 6
480
matrix-element calculations sufficiently describe the measured ratios,
indepen-481
dently of the Monte Carlo tune or parton-shower implementation.
482
The shape of the differential cross-sections as a function of transverse
mo-483
mentum and the scalar sum of transverse momenta, studied in the inclusive
484
two-jet and three-jet bins, drops off less (more) steeply in the 2 → n (2 → 2)
485
calculations. A measurement of the three-to-two-jet cross-section ratio as a
486
function of the leading jet transverse momentum and the sum of the two
lead-487
ing jet transverse momenta is better described by multi-leg parton programs.
488
Three-to-two differential cross-section ratios are compared to perturbative
next-489
to-leading-order QCD theoretical predictions, showing a reasonable agreement
490
with data.
491
Future comparisons with next-to-leading-order QCD calculations will
pro-492
vide useful information for constraining parameters, such as parton distribution
493
functions or the value of the strong coupling constant, αS. Systematic
un-494
certainties from the measurement are presently comparable to the theoretical
495
uncertainties, but should be reduced with larger data samples and better control
496
of all sources of systematic uncertainty.
Chapter 2
498
The Large Hadron Collider
4992.1
A Short Introduction to Lhc
500
The Large Hadron Collider (Lhc) at Cern near Geneva is the world’s newest
501
and most powerful tool for Particle Physics research [15, 16, 17]. It is designed
502
to collide proton beams with a center-of-mass energy of√s = 14 TeV and an
503
unprecedented luminosity of 1034cm−2s−1, a factor of ∼ 7 in energy and 100 in
504
luminosity larger than Tevatron at Fermilab. It can also collide heavy (P b) ions
505
with an energy of 2.8 TeV per nucleon and a peak luminosity of 1027cm−2s−1.
506
The colliding beam can contain up to 2808 proton bunches with up to 1.1 ×
507
1011protons each. The bunch-crossing rate will be 40 MHz at an instantaneous
508
luminosity of Linst = 2 × 1033cm−2s−1 in the low luminosity phase and Linst =
509
1034cm−2s−1 in the high luminosity phase.
510
The limiting factor to the achievable center-of-mass energy is the bending
511
power needed to keep the proton beams circulating in the 27 km-circumference
512
of the LEP tunnel. From the equation
513
p(TeV) ≃ 0.3B(T)R(km) (2.1)
where p is the beam momentum, B the magnetic filed provided by the magnets
514
of the accelerating machine and R ≃ 4.3 km is the radius of the Lhc ring, it is
515
deduced that the required magnetic field strength to achieve a beam momentum
516
of 7 TeV is about 5.4 T. In practice, since the machine cannot be completely
517
filled with magnets, the needed bending power is obtained with about 1200
su-518
perconducting dipoles providing a maximum magnetic field of 8.4 T at cryogenic
519
temperatures, which represents a very ambitious technological challenge.
520
The total inelastic proton-proton (pp) cross-section is approximately 80 mb
521
at √s = 14 TeV, meaning that Lhc will produce collision events at very high
522
rates. In particular, the event rate R, defined as the number of events produced
523
per second by the pp interactions, is expected to be
524
R = σ × L ≃ 80 mb × 1034 cm−2s−1≃ 109events/s (2.2)
when running at high luminosity.
525
On average 5 pp (low luminosity phase) and 25 pp (high luminosity phase)
526
interactions are expected per bunch collision. The most important nominal
527
parameters for the collider are summarized in Table 2.1.
528
Quantity Value
Circumference 26659 m
Dipole operating temperature 1.9 K
Number of magnets 9593
Number of main dipoles 1232
Number of main quadrupoles 392
Number of RF cavities 8 per beam
Nominal energy, protons 7 TeV
Nominal energy, ions 2.76 TeV per nucleon
Peak magnetic dipole field 8.33 T
Min. distance between bunches 7 m
Design luminosity 1034cm−2s−1
No. of bunches per proton beam 2808
No. of protons per bunch 1.15 × 1011
Number of turns per second 11245
Number of collisions per second 600 × 106
Stored energy per beam during collisions 362 MJ
Events per bunch crossing 19
Inelastic cross-section 60.0 mb
Total cross-section 80 − 100.0 mb
Table 2.1: Nominal parameters of the Lhc.
2.1. A SHORT INTRODUCTION TO LHC 9 Figure 2.1 shows an overview of the accelerator complex at Cern.
529
The protons are accelerated in several steps in the already existing
accelera-530
tor facilities. The protons originate from a hydrogen source and are accelerated
531
in the Linear Accelerator (LINAC) to an energy of 5 MeV. In the synchrotron
532
booster the protons obtain an energy of 1.4 GeV, then they are transferred into
533
the Proton Synchrotron (PS) and accelerated to 25 GeV further and finally
534
the energy is increased to 450 GeV in the Super Proton Synchrotron (SPS).
535
From the SPS they are injected into the Lhc ring where they are eventually
536
accelerated in RF-cavities up to their final energy.
537
For the acceleration there are eight RF-cavities installed per beam. Each
538
cavity provides an acceleration voltage of 2 MV at an operation frequency of
539
400 MHz. The bunch spacing in time is 25 ns. The cavities operate at a
540
temperature of 4.5 K. The two beams are accelerated in opposite directions
541
and have separate magnetic channels in the superconducting dipole magnets.
542
This is achieved by two apertures and an eight-shaped magnetic field. Both
543
beams share the same yoke and cryostat system.
544
In total there are 9593 installed superconducting magnets of which 1232 are
545
dipole magnets, 500 are quadrupole and 4000 are corrector magnets. The dipole
546
magnets are made of copper-clad niobium-titanium cables. With a total current
547
in the dipoles of 11.7 kA, a peak field of 8.33 T can be reached. The magnets
548
are cooled to a temperature of 1.9 K with super-fluid helium in a vacuum-vessel
549
contained cryostat. The beams travel in a beam pipe, which is held at a pressure
550
of 10−13 atm. In the collision areas the beam pipe is made of beryllium. The
551
beams are brought to collision at four points along the ring at which detectors
552
are positioned. Atlas and Cms are two multiple-purpose detectors that cover
553
a broad range of physics, whereas the Alice detector aims at heavy-ion physics
554
and LHCb at B-physics [17].
555
The Lhc construction was completed in 2008 and started the same year with
556
first beams. Due to problems with the dipole magnets at currents needed for
557
7 TeV-beams, the initial center-of-mass energy was planned to be√s = 10 TeV.
558
However, an incident in September 2008 with a superconductive connection bar
559
between two dipole magnets stopped the further running of the Lhc. This
560
connection bar inside the helium vessel had a small resistance that lead to
561
an electrical arc and it destroyed the helium enclosure. The sudden escape and
562
expansion of helium into the tunnel to atmospheric pressure lead to a mechanical
563
shock wave, which dislocated and damaged 58 of the dipole magnets. This
564
incident lead to a further decrease of the center-of-mass energy to a safer level
565
of √s = 7 TeV. After several months of repair, in November 2009 first pp
566
collisions were achieved with collision energies of up to√s = 2.36 TeV. In early
567
2010, the machine restarted and operating normally with beams and collisions
568
for a physics run with a collision energy of√s = 7 TeV.
569
By November 2011, Lhc has achieved the following operational records
570
Peak Stable Luminosity Delivered: 3.65 × 10
33cm−2s−1
571
Maximum Luminosity Delivered to ATLAS in one fill : 122.44 pb−1
572
Maximum Colliding Bunches: 1854
573
Maximum Peak Events per Bunch Crossing: 33.96
574
Maximum Average Events per Bunch Crossing: 32.21
By the end 2012 the √s = 7 TeV phase will be completed and a one year
576
shutdown with possible upgrades of the detectors will follow. In 2012 it is
577
foreseen to reach the design beam energies and design luminosity.
578
2.2
Impediments to High Luminosity
579
Hadron colliders employ bunched beams [18]. If two bunches containing n1and
580
n2 particles collide head-on with frequency f , instantaneous luminosity L is
581 given as 582 L = f4πσn1n2 xσy (2.3)
where σx and σy characterize the transverse beam profiles in the horizontal
583
(bend) and vertical directions. In this form, it is assumed that the bunches are
584
identical in transverse profile, that the profiles are independent of position along
585
the bunch and the particle distributions are not altered during bunch passage.
586
The single particle transverse motion in a alternating-gradient synchrotron
587
like Lhc, is not a simple sinusoid and rather it may be expressed in the form
588
x(s) = Apβ(s) cos[ψ(s) + δ] (2.4)
where s is path length in the beam direction, A and δ are constants of integration
589
and the envelope of the motion is modulated by the amplitude function, β. The
590
phase advances according to dψ/ds = 1/β; that is, β also plays the role of a
591
local wavelength λ/2π, and the tune ν is the number of such oscillations per
592
turn about the closed path. In the neighborhood of an interaction point, the
593
beam optics of the ring is configured so as to produce a near focus. The value
594
of the amplitude function at this point is designated β∗.
595
The motion as it develops with s describes an ellipse in {x, dx/ds}
phase-596
space, the area of which is πA2. If the interior of that ellipse is populated by
597
an ensemble of particles, given the parameter “emittance” and denoted by ǫ,
598
the area would change only with beam energy in the absence of other processes.
599
For a beam with a Gaussian distribution in {x, dx/ds}, the area containing one
600
standard deviation σx is the definition of emittance
601 ǫx= π σ2 x βx (2.5) with a corresponding expression in the other transverse direction, y. This
defi-602
nition includes ∼ 40% of the beam.
603
To complete the coordinates used to describe the motion, the longitudinal
604
variables {z, δp/p} is added in the transverse phase-space {x, dx/ds, y, dy/ds},
605
where z is the distance by which the particle leads the “ideal” particle along
606
the design trajectory. Radiofrequency electric fields in the s direction provide
607
the means for longitudinal oscillations, and the frequency determines the bunch
608
length. The quantity δp/p is characterized as “energy spread”.
609
In hadron collisions, the “bunch length” is a significant quantity for a
va-610
riety of reasons. If the bunch length becomes larger than β∗ the luminosity is
611
adversely affected. This is because β grows parabolically as one proceeds away
612
from the interaction point and so the beam size increases thus lowering the
613
contribution to the luminosity from such locations.
2.2. IMPEDIMENTS TO HIGH LUMINOSITY 11 Equation 2.3 can be recast in terms of emittance and amplitude functions
615 as 616 L = f4p n1n2 ǫxβx∗ǫyβy∗ (2.6) Therefore, to achieve high luminosity, all one has to do is make high
popula-617
tion bunches of low emittance to collide at high frequency at locations where
618
the beam optics provides as low values of the amplitude functions as possible.
619
While there are no fundamental limits to this process, there are certainly several
620
technical challenges to meet, mostly related to the structure of the acceleration
621
machine.
Chapter 3
623
Overview of the ATLAS
624Detector
625The Atlas detector is one of the two general purpose detectors constructed for
626
the Large Hadron Collider at Cern in Geneva. The detector is designed to be
627
sensitive to the full range of high pT physics processes occurring in√s = 14 TeV
628
proton-proton collisions and at high luminosity of 1034 cm−2s−1.
629
The hermetic Atlas detector, depicted by Figure 3.1, is composed of a
cen-630
tral tracker, a calorimeter system (electromagnetic and hadronic) and of a large
631
muon spectrometer. A detailed description of the detector is well documented
632
at Refs.[19, 20, 21, 22].
Figure 3.1: Cut-away view of the Atlas detector. The dimensions of the detector
are ∼ 25 m in height and ∼ 44 m in length. The overall weight of the detector is approximately 7000 tonnes.
633
The physics program of Atlas can be principally sorted into four categories.
634
The first covers the SM and flavor physics, Higgs explorations, searches for
635
η-coverage
Detector Component Required Resolution Measurement L1 Trigger
Tracking σpT pT = 0.05%pT ⊕ 1% ±2.5 -EM calorimetry σE E = 10% √ E ⊕ 0.7% ±3.2 ±2.5 Hadronic calorimetry
Barrel & End-Cap σE
E = 50% √ E ⊕ 3% ±3.2 Forward σE E = 100% √ E ⊕ 10% 3.1 < |η| < 4.9 Muon spectrometer σpT pT = 10% ±2.7 ±2.4
Table 3.1: Atlas design performance requirements [22]. The Muon spectrometer
performance is quoted for a muon with pT = 1 TeV, measured in stand-alone mode,
independently of the Inner Detector.
physics beyond the SM and new phenomena like exotic physics and gravitation
636
at tera-scales. In the second category belongs the physics of heavy ion collisions.
637
To accommodate this rich physics program, Atlas has a general-purpose design,
638
capable of detecting and measuring different types of particles. The main design
639
aims of Atlas can be summarized as follows [19, 20]:
640
Fast, radiation-hard electronics and sensors with high granularity;
641
Excellent momentum resolution, detector efficiency and vertex
identifica-642
tion;
643
Electromagnetic calorimetry for electron and photon measurements, and
644
full coverage hadronic calorimetry for jet and missing ET measurements;
645
Muon identification and measurement over a wide range of energies;
646
High efficiency triggers with excellent background rejection, capable of
647
working with low kinematic thresholds in a high multiplicity environment;
648
High energy/momentum resolution in all sub-detectors, summarized in
649
Table 3.1.
650
All of this must be achieved with high precision in the challenging
environ-651
ment set by the Lhc machine, with up to an anticipated average number of 23
652
inelastic pp collisions per 25 ns bunch crossing, consisting mostly of inelastic
653
QCD processes.
654
Emphasis is given on efficient tracking identification of charged particles and
655
accurate, large acceptance calorimetric measurement of shower pT and EmissT .
656
3.1
Kinematic Definitions
657
Throughout the following descriptions of the Atlas detector, cylindrical
co-658
ordinates R and φ are mostly used in the transverse plane x − y. In Atlas
659
detector, the positive x-axis is defined as pointing from the interaction point to
660
the center of the LHC ring, the positive y-axis is defined as pointing vertically
661
upwards, and the positive z-axis corresponds to protons running anti-clockwise.
3.1. KINEMATIC DEFINITIONS 15 The polar angle θ is measured from the beam axis (z-axis), the azimuthal angle
663
φ is measured in the transverse x − y-plane.
664
In experimental particle physics, a convenient set of variables to describe the
665
kinematics of measured objects is the Lorentz-invariant pseudorapidity η and
666
transverse component of momentum pT and the azimuthal angle φ.
667
Instead of polar angle θ, pseudorapidity is a commonly used spatial
coordi-668
nate describing the angular direction of a particle relative to the beam axis. It
669 is defined as 670 η = − ln tanθ 2 (3.1) where θ is the angle between the particle momentum vector p and the beam
671
axis. In terms of the momentum p = (px, py, pz), the pseudorapidity variable
672 can be written as 673 η = 1 2ln |p| + pL |p| − pL (3.2)
where pL= pZ is the component of the momentum along the beam axis.
There-674
fore, pseudorapidity depends only on the polar angle of the object’s trajectory,
675
and not on its energy. It is straightforward to express η as a function of rapidity
676
y and vice versa. From the definition of η, we have
677 eη= s |p| + pz |p| − pz (3.3) and 678 e−η= s |p| − pz |p| + pz (3.4) which combined together give
679
|p| = pTcosh η (3.5)
where pT is the magnitude of the transverse momentum
680
pT =
p
p2− p2
z. (3.6)
Subtracting Equation 3.4 from Equation 3.3, we obtain
681
pz= pTsinh η. (3.7)
Using these results, one can define the rapidity variable y in terms of the
pseu-682
dorapidity variable η as follows
683 y = 1 2ln q p2 Tcosh 2η + m2+ p Tsinh η q p2 Tcosh 2η + m2− p Tsinh η (3.8)
where m is the rest mass of the measured particle or object. Conversely, the
684
pseudorapidity η can be expressed in terms of the rapidity y by
685 η =1 2ln q p2 Tcosh 2 y − m2+ m Tsinh y q p2 Tcosh 2 y − m2− m Tsinh y (3.9)
where 686 mT = E cosh y = pz sinh y (3.10)
is conventionally called the “transverse mass” of particle with energy E, given
687
by
688
m2T = m2+ p2x+ p2y. (3.11)
So, rapidity y can be also defined as
689 y = 1 2ln E + pL E − pL = tanh−1 pL E (3.12) In the high relativistic limit, p ≫ m, the rapidity defined in Equation 3.8
690
may be expanded to approximately obtain
691 y ≃ − ln tanθ 2 ≡ η (3.13)
getting thus back the expression of Equation 3.1. The pseudorapidity η is
ap-692
proximately equal to the rapidity y for relativistic objects, p ≫ m and θ ≫ 1/γ,
693
and in any case can be measured in detectors when the mass and momentum of
694
the particle are unknown. From this definition, one can also obtain the identities
695
sinh η = cot θ, cosh η = 1/ sin θ, tanh η = cos θ. (3.14)
Often, one speaks of the “forward” direction in a hadron collider experiment,
696
which refers to regions of the detector that are close to the beam axis, at high
697
η. The difference in the pseudorapidity of two objects is independent of Lorentz
698
boosts along the longitudinal axis.
699
Te radial distances between two points in the η − φ plane are often denoted
700
by
701
R2= (δη)2+ (δφ)2. (3.15)
In some cases, it is more convenient to define transverse energy as the energy
702
deposited in a calorimeter component, corrected for its position by the formula
703
ET = E/ cosh η. (3.16)
In the highly relativistic limit, E ≫ m, and neglecting calorimeter resolution
704
effects, the deposited ET is equal to the pT of the incident particle.
705
3.2
ATLAS Magnets and Magnetic Field
706
Atlas features a unique hybrid system of four large superconducting magnets.
707
This magnetic system is 22 m in diameter and 26 m in length, with a stored
708
energy of 1.6 GJ. After approximately 15 years of design, construction in
indus-709
try, and system integration at Cern, the system was installed and operational
710
in the underground cavern in 2007.
711
Figure A.1 shows the general layout, the four main layers of detectors and the
712
four superconducting magnets which provide the magnetic field over a volume of
713
approximately 12000 m3(defined as the region in which the field exceeds 50 mT).
714
The Atlas magnet system, whose main parameters are listed in Table A.1,
715
consists of:
3.2. ATLAS MAGNETS AND MAGNETIC FIELD 17
a solenoid (Section 3.2.1), which is aligned on the beam axis and provides
717
a 2 T axial magnetic field for the inner detector, while minimizing the
718
radiative thickness in front of the barrel electromagnetic calorimeter;
719
a barrel toroid (Section 3.2.2) and two end-cap toroids (section 2.1.3),
720
which produce a toroidal magnetic field of approximately 0.5 T and 1 T
721
for the muon detectors in the central and end-cap regions, respectively.
722
3.2.1
Central Solenoid
723
The central solenoid [23, 24] is designed to provide a 2 T axial field (1.998 T at
724
the magnet’s center at the nominal operational current of 7.730 kA. To achieve
725
the desired calorimeter performance, the layout was carefully optimized to keep
726
the material thickness in front of the calorimeter as low as possible, resulting in
727
the solenoid assembly contributing a total of ∼ 0.66 radiation lengths at normal
728
incidence. This requires, in particular, that the solenoid windings and LAr
729
calorimeter share a common vacuum vessel, thereby eliminating two vacuum
730
walls. An additional heat shield consisting of 2 mm thick aluminium panels is
731
installed between the solenoid and the inner wall of the cryostat. The
single-732
layer coil is wound with a high-strength aluminium-stabilized N bT i conductor,
733
specially developed to achieve a high field while optimizing thickness, inside a
734
12 mm thick aluminium support cylinder.
735
The inner and outer diameters of the solenoid are 2.46 m and 2.56 m and
736
its axial length is 5.8 m. The coil mass is 5.4 tonnes and the stored energy is
737
40 MJ. The stored-energy-to-mass ratio of only 7.4 kJ/kg at nominal field clearly
738
demonstrates successful compliance with the design requirement of an extremely
739
light-weight structure. The flux is returned by the steel of the Atlas hadronic
740
calorimeter and its girder structure. The solenoid is charged and discharged in
741
about 30 minutes. In the case of a quench, the stored energy is absorbed by
742
the enthalpy of the cold mass which raises the cold mass temperature to a safe
743
value of 120 K maximum. Re-cooling to 4.5 K is achieved within lees than one
744
day.
745
The electromagnetic forces are counteracted by the combination of the coil
746
and warm-to-cold mechanical support, which maintains the concentricity of the
747
windings. All solenoid services pass through an S-shaped chimney at the top of
748
the cryostat, routing the service lines to the corresponding control dewar.
749
3.2.2
Toroid
750
The Atlas Toroid Magnet system consists of eight Barrel coils housed in
sepa-751
rate cryostats and two end-cap cryostats housing eight coils each. The end-cap
752
coils systems are rotated by 22.5◦ with respect to the Barrel Toroids in order
753
to provide radial overlap and to optimize the bending power in the interface
754
regions of both coil systems.
755
Barrel Toroid
756
The main parameters of the magnet are listed in Table A.1. The cylindrical
757
volume surrounding the calorimeters and both end-cap toroids (see Figure A.1)
758
is filled by the magnetic field of the barrel toroid, which consists of eight coils
assembled radially and symmetrically around the beam axis and encased in
in-760
dividual racetrack-shaped, stainless-steel vacuum vessels (see Figure A.2). The
761
coil assembly is supported by eight inner and eight outer rings of struts. The
762
coils are of a flat racetrack type with two double-pancake windings made of
763
20.5 kA Al-stabilized N bT i superconductor. Each coil has an axial length of
764
25.3 m and extends radially from 9.4 m to 20.1 m. The total assembly weights
765
about 830 tonnes. The peak field provided by the Barrel Toroid coils is 3.9 T,
766
providing 2 to 6 Tm of bending power in the pseudorapidity range from 0 to
767
1.3.
768
The conductor and coil-winding technology is essentially the same in the
769
barrel and end-cap toroids; it is based on winding a pure Al-stabilized N bT i/Cu
770
conductor into pancake-shaped coils, followed by vacuum impregnation.
771
The cool down of the 360-tonne cold mass to 4.6 K takes about five weeks.
772
The net Lorentz forces of approximately 1400 tonnes per coil directed inwards
773
and the self-weight of the toroids are counteracted by the warm structure of
774
Al-alloy struts mounted in between the eight coils.
775
End-cap Toroids
776
The Atlas end-cap Toroid systems (Figure A.3) consists of eight coils
assem-777
bled radially and symmetrically around the beam axis. The coils are of a flat
778
racetrack type with two double-pancake windings made of 20.5 kA Al-stabilized
779
N bT i superconductor. They are cold-linked and assembled as a single cold mass
780
in one large cryostat. The cryostat rests on a rail system facilitating the
move-781
ment and parking for access to the detector center. Each coil has an axial length
782
of 5 m and extends radially from 1.65 m to 10.7 m. The total assembly weights
783
about 239 tonnes. The peak field provided by the Barrel Toroid coils is 4.1 T,
784
providing 4 to 8 Tm of bending power in the pseudorapidity range from 1.6 to
785
2.7.
786
3.3
Tracking System
787
3.3.1
Overview of the ATLAS Inner Detector
788
The Atlas Inner Detector (ID) [25] is designed to provide hermetic and
ro-789
bust pattern recognition, excellent momentum resolution and both primary and
790
secondary vertex measurements for charged tracks above a given pT threshold
791
(nominally 0.5 GeV within the pseudorapidity range |η| < 2.5 and full azimuthal
792
coverage). It also provides electron identification over |η| < 2.0 and a wide range
793
of energies (between 0.5 GeV and 150 GeV). This performance is required at
794
very dense environments and at highest luminosities expected from pp collisions
795
at Lhc. The general ID layout, as shown in Figure 3.2, reflects the performance
796
requirements.
797
The detector has been designed to provide a transverse momentum
resolu-798
tion, in the plane perpendicular to the beam axis, of σpT/pT = 0.05%pT GeV ⊗
799
1% and a transverse impact parameter resolution of 10µm for high momentum
800
particles in the central η region.
801
The Atlas Inner Detector combines high-resolution detectors at the inner
802
radii with continuous tracking elements at the outer radii, all contained in the
3.3. TRACKING SYSTEM 19
Figure 3.2: A 3D model of the Atlas Inner Detector.
Central Solenoid, which provides a nominal field of 2 T. The highest
granu-804
larity is achieved around the vertex region using semiconductor pixel detectors
805
followed by a silicon micro-strip detector. Typically for each track the pixel
806
detector contributes three and the strips four space points. At larger radii
typ-807
ically 36 tracking points are provided by the straw tube tracker. The relative
808
precision of the measurement is well matched, so that no single measurement
809
dominates the momentum resolution. The outer radius of the Inner Detector
810
is 1.15 m, and the total length 7 m. In the barrel region the high-precision
811
detectors are arranged in concentric cylinders around the beam axis, while the
812
end-cap detectors are mounted on disks perpendicular to the beam axis. The
813
barrel TRT straws are parallel to the beam direction. All end-cap tracking
814
elements are located in planes perpendicular to the beam direction.
815
The Inner Detector comprises three complementary sub-detectors: the Pixel
816
Detector, the Semiconductor Tracker and the Transition Radiation Tracker.
817
Relevant features are described briefly below.
818
3.3.2
Pixel Detector
819
It consists of sensitive elements cover radial distances between 50.5 mm and 150
820
mm. The detector consists of 1744 silicon pixel modules [22] arranged in three
821
concentric barrel layers and two end-caps of three disks each. It provides
typi-822
cally three measurement points for particles originating in the beam-interaction
823
region. Each module covers an active area of 16.4 mm × 60.8 mm and contains
824
47232 pixels, most of size 50µm×400µm. The direction of the shorter pitch
de-825
fines the local x-coordinate on the module and corresponds to the high-precision
826
position measurement in the r − φ plane. The longer pitch, corresponding to
827
the local y-coordinate, is oriented approximately along the z direction in the
828
barrel and along r in the end-caps. A module is read out by 16 radiation-hard